Editing Learning (section)

== In animal evolution ==
Animals gain knowledge in two ways. First is learning—in which an animal gathers information about its environment and uses this information. For example, if an animal eats something that hurts its stomach, it learns not to eat that again. The second is innate knowledge that is genetically inherited. An example of this is when a horse is born and can immediately walk. The horse has not learned this behavior; it simply knows how to do it.<ref name="Dunlap"><{{cite thesis |author=Aimee Sue Dunlap-Lehtilä |url=http://conservancy.umn.edu/bitstream/51978/1/DunlapLehtila_umn_0130E_10349.pdf |title=Change and Reliability in the Evolution of Learning and Memory |access-date=2013-12-15 |type=PhD |publisher=University of Minnesota |archive-date=2013-11-13 |archive-url=https://web.archive.org/web/20131113233355/http://conservancy.umn.edu/bitstream/51978/1/DunlapLehtila_umn_0130E_10349.pdf}}></ref> In some scenarios, [[Instinct|innate knowledge]] is more beneficial than learned knowledge. However, in other scenarios the opposite is true—animals must learn certain behaviors when it is disadvantageous to have a specific innate behavior. In these situations, learning [[Evolution|evolves]] in the species.

=== Costs and benefits of learned and innate knowledge ===
In a changing environment, an animal must constantly gain new information to survive. However, in a stable environment, this same individual needs to gather the information it needs once, and then rely on it for the rest of its life. Therefore, different scenarios better suit either learning or innate knowledge.
Essentially, the cost of obtaining certain knowledge versus the benefit of already having it determines whether an animal evolved to learn in a given situation, or whether it innately knew the information. If the cost of gaining the knowledge outweighs the benefit of having it, then the animal does not evolve to learn in this scenario—but instead, non-learning evolves. However, if the benefit of having certain information outweighs the cost of obtaining it, then the animal is far more likely to evolve to have to learn this information.<ref name="Dunlap" />

Non-learning is more likely to evolve in two scenarios. If an environment is static and change does not or rarely occurs, then learning is simply unnecessary. Because there is no need for learning in this scenario—and because learning could prove disadvantageous due to the time it took to learn the information—non-learning evolves. Similarly, if an environment is in a constant state of change, learning is also disadvantageous, as anything learned is immediately irrelevant because of the changing environment.<ref name="Dunlap" /> The learned information no longer applies. Essentially, the animal would be just as successful if it took a guess as if it learned. In this situation, non-learning evolves. In fact, a study of ''[[Drosophila melanogaster]]'' showed that learning can actually lead to a decrease in productivity, possibly because egg-laying behaviors and decisions were impaired by interference from the memories gained from the newly learned materials or because of the cost of energy in learning.<ref>{{cite journal | doi = 10.1016/j.anbehav.2003.12.005 | volume=68 | issue=3 | title=An operating cost of learning in Drosophila melanogaster | journal=Animal Behaviour | pages=589–598| year=2004 | last1=Mery | first1=Frederic | last2=Kawecki | first2=Tadeusz J. | s2cid=53168227 | url=http://doc.rero.ch/record/4149/files/1_kawecki_ocl.pdf }}</ref>

However, in environments where change occurs within an animal's lifetime but is not constant, learning is more likely to evolve. Learning is beneficial in these scenarios because an animal can [[Adaptation|adapt]] to the new situation, but can still apply the knowledge that it learns for a somewhat extended period of time. Therefore, learning increases the chances of success as opposed to guessing.<ref name="Dunlap" /> An example of this is seen in aquatic environments with landscapes subject to change. In these environments, learning is favored because the fish are predisposed to learn the specific spatial cues where they live.<ref>{{cite journal | last1 = Odling-Smee | first1 = L. | last2 = Braithwaite | first2 = V.A. | year = 2003 | title = The role of learning in fish orientation | journal = Fish and Fisheries | volume = 4 | issue = 3| pages = 235–246 | doi = 10.1046/j.1467-2979.2003.00127.x | bibcode = 2003FiFi....4..235O}}</ref>